chenchenghao/sglang
1
0
Fork 0
Code Issues Pull Requests Actions 12 Packages Projects Releases Wiki Activity

ci: improve nightly-ci (#11385)

Browse Source
This commit is contained in:
Mick
2025-10-13 12:19:34 +08:00
committed by GitHub
parent a55cf5304a
commit 0c0779d667
6 changed files with 76 additions and 54 deletions
Download Patch File Download Diff File Expand all files Collapse all files

21
python/sglang/bench_serving.py
View File

@@ -12,7 +12,6 @@ python3 -m sglang.bench_serving --backend sglang --dataset-name random --num-pro
import argparse
import asyncio
import base64
import io
import json
import os
@@ -671,7 +670,7 @@ def get_processor(
if pretrained_model_name_or_path.endswith(
".json"
) or pretrained_model_name_or_path.endswith(".model"):
from sglang.srt.hf_transformers_utils import get_processor
from sglang.srt.utils.hf_transformers_utils import get_processor
return get_processor(pretrained_model_name_or_path)
@@ -935,7 +934,7 @@ async def get_mooncake_request_over_time(
for i in range(num_rounds):
# Add user query for the current round
chat_history.append(
{"role": "user", "content": f"Round {i+1}: {user_query_base}"}
{"role": "user", "content": f"Round {i + 1}: {user_query_base}"}
)
# Form the full prompt from history
@@ -964,7 +963,7 @@ async def get_mooncake_request_over_time(
def sample_mmmu_requests(
num_requests: int,
processor: AutoProcessor,
processor: AutoProcessor | AutoTokenizer,
fixed_output_len: Optional[int] = None,
random_sample: bool = True,
) -> List[DatasetRow]:
@@ -973,9 +972,7 @@ def sample_mmmu_requests(
Args:
num_requests: Number of requests to sample.
tokenizer: Tokenizer to use for token counting.
fixed_output_len: If provided, use this fixed output length for all requests.
apply_chat_template: Whether to apply the chat template to the prompt.
random_sample: Whether to randomly sample or take the first N.
Returns:
@@ -1282,11 +1279,11 @@ def parse_image_resolution(image_resolution: str) -> Tuple[int, int]:
)
def create_mm_data_row(text_prompt, images, images_base64, output_len, processor):
def create_mm_data_row(text_prompt, images: list, images_base64, output_len, processor):
try:
content_items = [
{"type": "image_url", "image_url": {"url": img_url}}
for img_url in images_base64
{"type": "image", "image": {"url": image_base64}}
for image_base64 in images_base64
]
content_items.append({"type": "text", "text": text_prompt})
prompt_str = processor.apply_chat_template(
@@ -1294,7 +1291,9 @@ def create_mm_data_row(text_prompt, images, images_base64, output_len, processor
add_generation_prompt=True,
tokenize=False,
)
except Exception:
except Exception as e:
# Note (Xinyuan): This is a workaround for an issue where some tokenizers do not support content as a list. (e.g. InternVL)
print(f"Error applying chat template: {e}, fallback to <image> tag")
# Some tokenizers do not support list content; fall back to a placeholder in the text
prompt_str = f"<image>{text_prompt}"
@@ -1425,7 +1424,7 @@ def sample_image_requests(
print(f"#Input tokens: {np.sum([x.prompt_len for x in dataset])}")
print(f"#Output tokens: {np.sum([x.output_len for x in dataset])}")
print(
f"\nCreated {len(dataset)} {image_content} {image_format} images with average {total_image_bytes//num_requests} bytes per request"
f"\nCreated {len(dataset)} {image_content} {image_format} images with average {total_image_bytes // num_requests} bytes per request"
)
return dataset